Successful inventory control depends on a combination of factors.

Timely and accurate data is the first factor. Inventory control is the discovery and implementation of the best answers to a few simple questions. What should be stocked? When should orders be placed? How much should be ordered? Simple on paper, these questions become complex when asked in the context of the reality of business. The reality of business - or at least, the reality of business as best our minds can represent it - comprises numerous energy and material flows as well as psychological factors operating on both the micro and macro scales. Timely and accurate data about demand, available supply, existing stock, and plant or warehouse capacity is the beginning of the process by which business reality becomes transformed into the figures and abstracted models necessary for forecasting and decision-making. Humanity succeeds on the strength of our ability to translate the infinitely various details of the world of phenomena into models that are both simple enough to hold in the mind and accurate enough to be a sound basis for advantageous decisions in a particular context.

Models are our second factor. An example from another field: Rivers are infinitely complex flows that have fascinated humanity for all of its history. The potential of rivers for fueling contemplation, poetry, reverie, and philosophy is inexhaustible. Likewise, the potential of rivers for creating unique human experiences is inexhaustible. But an engineer designing a dam can "reduce" the infinite potential of the river to those few numeric values that are relevant to the business of dam-building. This is what gives the engineer power over the river. The dam-building is the particular context which allows the engineer to focus attention on only a few factors: rate of water flow, depth and width of river, average seismic activity in the region, temperature variation, strength of building materials, and so on. The engineer's training and experience give him or her a mental model of dams and rivers into which objective factors can be plugged to yield proper building decisions.

Both the model and the data that goes into the model must be accurate for correct decisions to be made. Inaccurate data going into a perfect model yields inaccurate results: garbage in, garbage out. Perfect data going into an inaccurate model likewise yields inaccurate results: a driver with perfect vision cannot steer well if the steering mechanism is out of order.

In the inventory control world, simple and effective models have been available for more than fifty years. The last fifty years have also seen the permeation of business by the computer: a machine which has replaced the often makeshift, haphazard, or cumbersome inventories of the past with database systems that are reliable, centralized, and easy to use. One of the surprises of the last half-century has been the success of this world-changing technology that not only matched but exceeded most of the hype in its attendance, replacing file cabinets, punch-cards, and human memory with a plastic box on top of a desk.

The computer has also helped greatly enhance our third factor, which is forecasting power. Inventory control can be defined as the maintenance of the most efficient, least expensive buffer between demand and supply. The more one knows about future demand and future supply, the less safety stock one has to hold. An inventory controller who knew exactly what would be available when, and exactly what would be in demand when, could get by with minimal stock. Every improvement in predictive power directly enhances the efficiency of stock control.

But the accurate data upon which forecasting depends means something, as we have said before, only in a certain context. To know what to do with data one must know what one wishes to accomplish: the goal or objective. This objective, or system of objectives, is our fourth factor. Most stock-holding businesses describe their objective in terms of two figures: service level, which is the percentage of customer orders that can be met directly from the business' own stock, without backordering - and the cost of holding stock. Service level is to be maximized and the cost of holding stock is to be minimized. Both are objective figures that directly index the business' efficiency and profit - hence they are so valuable. With vague, unobjective, or mistaken goals it is easy for a manager to believe that stock control is improving when it is actually falling apart. Having precise objectives also gives employees and managers the valuable psychological advantage of being self-directed actors capable of exactly measuring their performance. (Just-in-Time is not an exception because in the Just-in-Time methodology goals are actually even more methodic and objective than in most traditional systems.)

Our fifth factor is the actual human, managerial and physical ability to implement the decisions of inventory control. This is the bottleneck factor without which the others are irrelevant, because it is the factor that turns theoretical inventory control into practical inventory control. It comprises everything between the making of the decision and its final, physical result, which can then be measured against set objectives.

These five factors - timely and accurate data, appropriate inventory control models, accurate forecasting, meaningful objectives, and implementation - are equally important because each depends on all of the others. Without data and forecasting there is nothing to feed into the model. Without objectives there is no way to measure success or failure. Without implementation inventory control is merely theoretical. An improvement in any of the factors is an improvement in the whole.